Heterogeneous Ensemble Learning for Context-Aware Image Captioning with Transformers. JSIR. 2026;84(12):1322-1330. doi:10.56042/jsir.v84i12.23517